Candidate-gene association studies for coronary heart disease (CHD) often suffer from variability and failure of replication. This may partly result from a failure to understand linkage disequilibrium (LD) within a gene and, thus, inappropriate selection of a genetic variant for disease association testing. Genomic fragments (haplotype blocks) exist that experience infrequent recombination events at the population level and exist as invariant blocks interspersed with "hotspots" of recombination. Ambiguous disease associations can result when a genetic marker is non-functional but is linked to a functional, disease-causing variant on one (or a few) haplotype(s), but not on others. It has been proposed that association studies using variants representing haplotype blocks may provide greater power, and, thus, better economy and efficiency. Haplotype blocks can be determined by creating a dense map of single nucleotide polymorphisms (SNPs) across the gene of interest and analyzing the population-level LD. A SNP (or a few SNPs) can then be chosen to represent each haplotype block, and used to assess disease association for the blocks across the entire gene. Elevated low-density lipoprotein (LDL) and reduced high-density lipoprotein (HDL) are risk factors for coronary artery disease (CAD). Genes regulating lipoprotein metabolism influence cholesterol levels and may provide the basis for genetic predisposition to CAD. The cholesteryl ester transfer protein (CETP) mediates transfer of cholesteryl esters from HDL to LDL and very-LDL, possibly shifting the ratio of these lipids towards a risk profile. Individual variants of the CETP gene have been investigated in CHD; however, these association studies are generally inconclusive or contradictory, possibly due to inadequate sample size or inappropriate marker variant selection, as proposed above. To date, no study has provided a characterization of the LD and haplotype structure of CETP SNPs or their association to disease endpoints. The objective of this study is to combine high-throughput genotyping capability with genetic epidemiological methods to identify the haplotype blocks within and surrounding the CETP gene and, using the currently largest CHD DNA bank with corresponding clinical data, to test CETP haplotype blocks for association to the following endpoints: (1) the diagnosis of angiographically-significant CAD, (2) response to therapy, (3) significant adverse CHD events (death, MI).